This proposal describes the development of a novel system for enabling high throughput mutation analysis of cancer. Currently, polymerase chain reaction (PCR) is a critical component for amplifying selected regions of the genome such as gene exons for subsequent resequencing. A major limitation of PCR is that amplification artifacts arise when large numbers of specific primer pairs are simultaneously added to a reaction. We have developed a solution to this problem that enables multiplex PCR amplification of specific target sequences without amplification artifacts. The procedure is based on oligonucleotide constructs, called selectors. The selectors identify defined target nucleic acid sequences, and they act as ligation templates to direct circularization of these targets. The selectors contain a general primer-pair motif that allows the circularized targets to be amplified in highly multiplexed reactions using a universal PCR primer pair. As part of the bioinformatics for choosing selector sequences, we developed a computer program called PieceMaker that finds the optimal sequence of selector probes for a given selector application. In this application, we propose to adapt selector technology with massively parallel sequencing systems to enable sequencing of large number of cancer genes. As a test case, our technology development is focused on developing a selector assay to sequence in parallel ten cancer genes in a single reaction volume. We will use a highly characterized group of over fifty colorectal carcinoma cell lines, many of them having identified mutations such as the TP53 gene. These cell lines will be used to validate the development of selector technology, optimize its ability to detect missense, insertion and deletion mutations and use known mutations to determine overall sensitivity and specificity. The accumulation of genetic errors, otherwise known as mutations, in specific genes contributes to the behavior of a cancer. Technologies to identify these mutations is limited and as a result, only a relatively few of these mutations have been identified in specific genes. We are developing a new technology that streamlines the process of identifying mutations such that large numbers of genes can be analyzed in parallel. With additional development, this technology will enable larger, more comprehensive identification of critical mutations in cancer and identification of these mutations have the potential to become important clinical tests. [unreadable] [unreadable] [unreadable]